Distributed databases may provide for storage of related data in multiple database devices. This may particularly be useful with data-intensive applications that store quantities of data which cannot typically be handled by a single database device. For example, it may be desirable to distribute data relating to geographical items (roads, intersections, individual addresses, etc.) for mapping applications among multiple database devices. As the number of defined geographical items desired to be stored may require a massive amount of database memory, distributing the geographical items across a distributed database comprising multiple database devices may be preferred. While generic distributed databases may be used to store such data, present systems may not provide for accurate results during parallel writes for transactions that implement changes across multiple database devices.